Save your Bugtracker/Mantis login => Reminder: Moving to GitLab

[PascalDragon]

I followed the instructions today, but it seams that deadline was yesterday. Did I really miss the chance?
https://bugs.freepascal.org/view.php?id=39246

We've pushed the conversion back a week due to some further issues.

Stonage IT

If you're so genius then how would you map users of one system to another without having any pre-existing mapping of the users between the two?

[Martin_fr]

I followed the instructions today, but it seams that deadline was yesterday. Did I really miss the chance?
https://bugs.freepascal.org/view.php?id=39246

Deadline extended. We are running a bit behind.
Deadline is, when mantis no longer lets you log in / or no longer exist.

Currently: ~23rd/24th/25th

[Fred vS]

how would you map users of one system to another without having any pre-existing mapping of the users between the two?

Mostly all people use the same user name for their Mantis-GitLab-GitHub account.
What about to try to set user name of Mantis for GitLab?

Of course there will be "unknown" users but also "known" users.

My 0.0001 cent, of course.

[Kays]

This is an arrogant attitude towards the people who do all the work for the move to git.

I second that.

If you're so genius then how would you map users of one system to another without having any pre-existing mapping of the users between the two?

I guess this excludes email addresses. At least some people have the same email address on GitLab as well as Mantis.

[…] So you need to create 2 issues, one in Mantis that ask the Gitlab account and one in GitLab that ask the Mantis account.
And after compare if both are the same.

You could probably eliminate one issue (at least the Mantis issue) if you simply demand a verifiable and trustworthy GPG-signed message using a key containing the email address [as an uid] as stored in the Mantis bug tracking system. However, trustworthy would imply a working (non-expired) web of trust link to one of the core developer’s key: problem is, are there any?

[Fred vS]

At least some people have the same email address on GitLab as well as Mantis.

Yep, well seen (I forgot the email).

Then first and most important, sent, by courtesy, a email to every Mantis user-account saying that fpc will go to git.  (Dont do like Nabble).
Also ask them if they have a gitlab account (I think that gitlab can deal with github account too).
If they have, answer to the mail with the gitlab account.

Also, maybe try to do a compare with email of user in GitLab and Mantis to get the respective user name.

[winni]

This is an arrogant attitude towards the people who do all the work for the move to git.

I was not asked if I like it to move to git.

Now I am told - hop. hop - to move to git.

THAT is arrogant.

Winni

[Martin_fr]

I was not asked if I like it to move to git.

Now I am told - hop. hop - to move to git.

THAT is arrogant.

<sarcasm>
No, arrogant would be the next line:

You don't need to, just download releases as before

</sarcasm>
SCNR

But seriously:

If you are a member of the core team, then yes: Everyone in the team moves.

Otherwise:
Unless you want to (or do) contribute, you do not *need* to move to git.
You can download any revision as zip or tar.gz directly for gitlab.

And even if you want to contribute, you can send patches.

If you want to report bugs, you need to register at gitlab, but not move to git.
The same as in the old time you had to register on mantis.

You also do not need to provide your gitlab name for existing bugs.
They will be moved, and you can search for them by your mantis username.

The ability to provide a gitlab account now, is merely a courtesy. If you wish that name to be listed in the converted bug reports that you made (or notes you added).

So really no "hop hop"... ?

Oh, and I have been told, but I have no idea if it true, or will indeed work: Github allows svn access.
We may (eventually) have an github mirror (read only). So if what I was told is true....

I still don't get how the above quoted statement of yours, relates to "stonage it"?

[PascalDragon]

how would you map users of one system to another without having any pre-existing mapping of the users between the two?

Mostly all people use the same user name for their Mantis-GitLab-GitHub account.
What about to try to set user name of Mantis for GitLab?

Of course there will be "unknown" users but also "known" users.

That would still leave users that wouldn't be correctly mapped. There are also users that have multiple accounts on Mantis (mainly due to some hiccups with Mantis itself), among them also the one or other core dev. Doing it this way we also essentially get an overview who is still active.

This is an arrogant attitude towards the people who do all the work for the move to git.

I was not asked if I like it to move to git.

Of course not, you're not a member of the development teams. This is about what the development teams want to use to further development of FPC and Lazarus. For the users it's simple the following (to translate a German saying): "eat it or die". This is just like it has always been: When you came here for the first time you had to adhere to SVN and Mantis. Now you simply have to adhere to GitLab. It's always the project (and its devs) that makes the rules, not the users.

[Fred vS]

It's always the project (and its devs) that makes the rules, not the users.

Hum not sure to adhere with this but yes, it is (sadly) how lot of things work.
But not always, sometimes it is the users that define the rules (it is the idea of "marketing").

Imho, the users should be the kings.

Fre;D

[Fred vS]

It's always the project (and its devs) that makes the rules, not the users.

Hum not sure to adhere with this but yes, it is (sadly) how lot of things work.
But not always, sometimes it is the users that define the rules (it is the idea of "marketing").

Imho, the users should be the kings.

Fre;D

Please ban me after this post if you want.

Sven, I dont like how you treat Winni.

Winni, like other precious users, helps a lot for fpc, and maybe much more than some pretentious "core" developers.

Fre;D

[howardpc]

Sven, I dont like how you threat Winni.

Fred, I fail to see what you find objectionable. I presume by "threat" you meant "treat" and not "threaten", but in either case I think you are over-reacting.


To point out the difference between a user (however helpful he is) and a member of the development team is merely to state a fact. It in no way diminishes his helpfulness or represents a slight on winni. His many numerous helpful contributions are valued, but not relevant to decisions that are taken by those who are ultimately responsible for the projects.

[Martin_fr]

Hum not sure to adhere with this but yes, it is (sadly) how lot of things work.
But not always, sometimes it is the users that define the rules (it is the idea of "marketing").

Imho, the users should be the kings.

If the users pay the developers, then the developers need to (or must) do whatever it takes to earn there money.

If the developers volunteer there free time, and give away there work for free, then they can choose.
That said, from my observation its not black or white. All the developers look at feedback, and look at were they can help others.

 

[Martin_fr]

Please ban me after this post if you want.

Why?
We allow and invite critics.

Same for Winni. His post is fine (though it was not immediately clear what it was about).

We may not be able to help him in this particular instance. Or at least not by reverting our decision.
And maybe this individual complaint is not going to change things for the future.

But that is not to say, that every complaint, critic or opinion is pointless. They are read. And maybe some are considered.

Just for info, at the moment I guess that winni's post is only about the way the account is transferred to gitlab?

Now I am told - hop. hop - to move to git.

That was in the context of this thread. The subject of it, he himself has made the "Save your mantis login".

And the related request that we posted, is indeed the one that says "please to so by deadline" (hence "hop hop"?)

But maybe he also means having to move from svn to git. Not 100% sure at this point.
(But even on this, answers have been given, attempting to provide help)

[devEric69]

Now I am told - hop. hop - to move to git.

That's for sure: a routing problem, amho.

[valdir.marcos]

@Winni

I agree with you that using issues (from Mantis or GitLab) is archaic.
And it will need lot of boring work for the maintainer to do manually the transfer for each answer.

Are you thinking to a special program to do the job, via internet and using database?

Mantis is a database frontend.

https://gitlab.com/freepascal.org/mantis_to_gitlab

Just curiosity.
Let's suppose you want to move this data to any other solution in the future.
Gitlab allows that as Mantis?

Of course!

For example I have one account in GitLab and one in GitHub.
All my GitHub repositories are mirrored and updated in (nearly) real time in the GitLab repositories.

And vice-versa, you may do the same from GitLab to GitHub.

Also I may login in GitLab with a GitHub account!

[EDIT] In "Settings-General" you have even a option to easy export the project (see picture):

What is "container registry images" that would not be exported?